Node.js Trading Bot

I run a Node.js trading bot that pulls market data from CoinGecko, feeds it into a lightweight AI decision layer, and then places orders through the exchange API. The logic is sound, yet the application behaves inconsistently,

My main goal is to eliminate these performance bottlenecks so the bot can operate around the clock without throwing errors. This involves tracing what happens during the external API calls, tightening up any throttling or retry logic, and putting rock-solid error handling in place so a single failure never stops the entire strategy. Along the way I expect best-practice improvements—structured logging, graceful shutdown hooks, and lightweight monitoring—to help me keep an eye on latency, memory, and CPU in production.

Deliverables
• Profiled, refactored code in Node.js that completes each CoinGecko (and similar) request within an acceptable, documented latency window
• Robust error-handling layer covering network faults, rate-limit responses, and AI inference exceptions
• A short README explaining any new environment variables, install steps, and how to read the performance metrics you add